Introduction

Apparent diffusion coefficient (ADC) has been widely established as a diagnostic biomarker. Some novel biomarkers has emerged from the fact that ADC depends strongly on the diffusion time, such as the rate of ADC change using two diffusion times for the differentiation of breast tumor types¹ and a cell size modeling technique call IMPULSED using three diffusion times². These novel techniques typically use a combination of oscillating gradient spin echo (OGSE) and pulsed gradient spin echo (PGSE) sequences for data acquisition^1,2, which usually requires a long TR, hence long scan time for the acquisition of multiple b values, diffusion times and large spatial coverage. In this study, we propose to use Multiband (MB) SENSE to accelerate the data acquisition and compare the diagnostic value of ADC from different diffusion times for breast tumor.

Methods

Data acquisition
Both OGSE and PGSE sequences were implemented and data were acquired on a Philips 3T scanner (Ingenia CX, Best, The Netherlands). 34 consecutive female patients (mean age, 50.18years) suspected of breast tumor/cancer were recruited between Feb 2021 to Oct 2021 and 34tumors were analyzed (18 benign and 16 malignant). The study was approved by the IRB of of People's Hospital of Ningxia Hui Autonomous Region and written informed consent was obtained from all the patients. The breast imaging protocol consisted of routine sequences (T1w, T2w, DCE) and an IMPULSED protocol, using MB=2. The sequence parameters for the IMPULSED protocol were listed in Table 1. Two OGSE sequences of 50Hz and 25 Hz and one PGSE sequence was used as reported in a previous study(2) (the corresponding diffusion time = 5 ms, 10 ms and 74 ms respectively). FOV =340×255mm2, in-plane resolution = 3*3 mm2, 16 slices with 5 mm slice thickness and 1 mm gap.
Data analysis
The ADC spectrum ADC(b, Δ) was calculated for each b value and diffusion time. 3 regions of interest (ROIs) were drawn on each lesion by a radiologist with 3 years of experience, referring to the DCE images, avoiding cyst and necrosis regions. The diagnostic performances to differentiate malignant and benign breast tumors for ADC calculated using b=0 and the largest b value for each sequence at the three diffusion times independently were investigated using the areas under the receiver operating characteristic curves (AUCs). The diagnostic performance using the combination of the ADC from three diffusion times was also investigated. The cutoff values were determined using Youden index.

Results

The ROI placement and mean ADC of one patient was shown in Figure 2, The AUC of ADC_5ms, ADC_10ms, ADC_74ms and the combination of the three were 0.91, 0.84, 0.76, and 0.90 respectively. The sensitivity and specificity and the cutoff values were listed in Table 2.

Discussion and conclusion

The ADC calculated using the shortest diffusion time showed the highest diagnostic performance, slightly higher than using three ADCs jointly. Although the effect of b value should also be accounted for, the largest achievable diffusion weighting for different diffusion times was restricted by the gradient performance of the system, especially for OGSE with high frequency. Specifically, the highest b value for OGSE at 50 Hz was about 300 s/mm2 on our current system and sequence implementation. However, for PGSE with a long diffusion time, the achievable b value was much higher. Our data showed superior diagnostic performance between benign and malignant breast lesions using ADC at a short diffusion time. Although further analysis should include the whole ADC spectrum to utilize the effect of diffusion weighting as well.

Acknowledgements

No acknowledgement found.